In: Statistics and Probability
A random sample of 100 observations yields a confidence interval with a margin of error of 60. In a new study we would like to estimate the same population parameter with a margin of error of 30 (half of the old margin of error). How many observations should we sample for this new study?
The margin of error will be half of the old margin of error, if we will take 4 times the old sample size.
So the required sample size = 100 * 4 = 400